PROJECT SUMMARY/ABSTRACT The discovery of ERG fusions in prostate cancer has set a paradigm where aberrant expression of non- mutated transcription factors at levels that are physiologic in other tissues can drive oncogenesis. We have uncovered that a subset of prostate cancers aberrantly express a gene signature normally restricted to the gastrointestinal (GI) tract and we call this the PCa-GI signature. The prevalence of this phenotype jumps from ~8% of primary prostate cancer to 30% of castration-resistance metastatic prostate cancer. Our data indicates that these GI genes are coordinately regulated by hepatic nuclear factor 1? (HNF1A) and hepatic nuclear factor 4? (HNF4G). These two transcription factor families (HNF1 and HNF4), together with any FOXA-family transcription factor, have been well characterized to form a core autoregulatory loop to govern the GI lineage specification and can reprogram fibroblasts into the GI lineage, similar to the ?Yamanaka factors? in governing the embryonic stem cell lineage. Prostate lineage express high endogenous levels of FOXA1. Our preliminary data using the 22Rv1 cell line that express the PCa-GI signature indicates that the HNF1A/HNF4G transcription circuit is required both to maintain expression of GI specific genes and for growth of HNF1A/HNF4G-positive prostate cancer cells. Further, ectopic expression of HNF4G in HNF1A/HNF4G- negative prostate cancer cells turns on the PCa-GI signature and leads to more rapid progression to castration-resistance. ChIP-seq studies show that HNF4G is necessary and sufficient to maintain GI lineage- specific enhancers, implying that HNF4G is a ?pioneer? transcription factor that can bind to closed chromatin to establish novel enhancers in the prostate lineage. The overall objective of our proposal is to understand the mechanistic role of the aberrant expression of HNF1A and HNF4G in prostate cancer tumorigenesis and progression to castration resistance. We will utilize next-generation patient-derived prostate cancer organoid models that are molecularly and clinical well annotated to define the broad requirement of the HNF1A/HNF4G circuit in tumors that aberrantly activates the PCa-GI signature. To understand their role in tumorigenesis, we will model ectopic HNF1A or HNF4G expression in a mouse prostate organoids isolated from genetically engineered mice with different combinations of SPOPF133V mutation, Chd1 loss, and Pten loss. To study their role in castration-resistance, we will dissect their interaction with AR-dependent transcriptome in both in vitro and in vivo models. We will further correlate gene expression with cistrome and chromatin landscape studies. If successful, our studies will define a novel mechanism of prostate cancer tumorigenesis and castration resistance. HNF1A/HNF4G can be developed as biomarkers. Furthermore, because HNF4G is a ligand dependent nuclear transcription factor, this subset of prostate cancer can potentially be therapeutically targeted.